Detection of Spam in YouTube Comments Using Different Classifiers

Abstract

Nowadays thousands of videos are uploaded to YouTube each minute and people start giving likes and comments immediately. Some popular and viral videos get millions of comments, some are healthy with good complements and some are Spams with unrelated, abusive, sometimes provided with a URL for commercial advertisement or redirection to other sites. In this paper, we have taken YouTube comment datasets of five famous singers and detecting Spam comments using some Artificial Neural Network based Classifiers and some Normal Classifiers. The proposed technique compares the derived results of the classifiers and suggests the best classifiers for detecting Spam comments.